Perplexity is doing something most people thought was impossible: competing with Google in search. At the Berkeley Haas AI Conference, their VP of Growth and Product, Rohan, laid out how they're doing it. The answer isn't a better algorithm or more data. It's speed.
The Mode Is Speed
Perplexity's CEO has said publicly that the company's mode is speed. Rohan confirmed this isn't just a talking point — it's the operating system of the entire company. "We live and die by just how fast we can move."
Their competitive logic is straightforward: if you're building something genuinely new — a new category of how people search and access information — your product IS your marketing. As long as you can keep shipping products that are fundamentally different and fundamentally better, you'll keep acquiring users. The moment you slow down, competitors catch up.
The question they ask themselves every single day: are we staying ahead of everyone by at least a couple of months?
Three Pillars of Speed
Rohan broke their approach down into three components:
1. Hire for dopamine.
They specifically hire people who "get their dopamine hits off of speed and being able to ship and see impact as quickly as possible." This isn't about work ethic — it's about wiring. Some people light up when they ship something on Monday and see the data on Tuesday. Those are Perplexity people.
2. Develop killer intuition about focus.
Speed without focus is just thrashing. The hardest part is knowing when to go all-in versus when to wait and see. "To a certain extent, it's a little bit of guessing and you hope your intuition gets you there." This is an underappreciated skill — the ability to make fast, high-conviction bets about where to invest effort.
3. Empower every layer.
Perplexity runs an incredibly flat organization. There is no hierarchy of "should I do this?" The expectation for every person: you are in charge of this. You are going to move these metrics. You are going to drive this as aggressively as you can. No permission-seeking. No committee approvals.
The Experimentation Machine
The section on experimentation was the most tactically rich part of the conversation. Perplexity runs what Rohan called "an endless amount of experimentation" on how they answer questions.
The fundamental insight: the right answer format varies wildly by query type. If you search for a cooking recipe, you actually want a very long answer — almost a blog post. Other queries want concise, direct responses. The only way to discover these patterns is relentless A/B testing.
Another surprising finding: onboarding length. Should you walk new users through the product, or just drop them in? After extensive testing, the winning approach has been to mostly just let people figure it out. Minimal hand-holding. Trust the product to be intuitive enough.
"That's something we're still tinkering with," Rohan admitted. But the willingness to keep tinkering, to never consider any part of the experience settled, is the point.
Hiring Underdogs
When asked what qualities Perplexity looks for when hiring, Rohan's answer was surprising: underdogs.
"Who's the underdog? Who's going to take big risks? We hire a lot of founders. Have they tried to spin something up of their own and failed gloriously? That's the kind of stuff where we're like — dang, check mark, get him in here."
They also rarely hire for defined roles. Instead, they hire for "where the puck is going" — deliberately ambiguous positions that expand rapidly as the company grows. They hire overqualified people for seemingly small roles, expecting those roles to balloon.
Why This Matters Beyond Perplexity
You don't have to be competing with Google for the Perplexity playbook to be relevant. The underlying principles apply to any AI product:
- **Your product velocity is your moat.** In a market where models improve monthly and competitors can clone features in weeks, the only sustainable advantage is shipping faster than everyone else.
- Flat orgs win in AI. Permission hierarchies are a speed tax. Every layer of approval is time your competitors are using to ship.
- Hire for energy, not just experience. People who viscerally enjoy shipping fast will outperform people who are merely competent but need deliberation cycles.
- Experiment on everything. The intuitive answer is wrong more often than you think. Test aggressively, even on things that seem settled.
Perplexity is proof that speed, combined with taste and relentless experimentation, can take on incumbents with 1000x more resources. In the AI era, being small and fast beats being big and slow every time.